| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 280, Issue 19, 19185-19195, May 13, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
-Catenin Signals Is Required for Growth Plate Assembly, Cartilage Integrity, and Endochondral Ossification*
From the
Department of Orthopaedic Surgery, Jefferson Medical College, Thomas Jefferson University, Philadelphia, Pennsylvania 19107, the Department of Oral Cytology and Cell Biology, School of Dentistry, Nagasaki University, Nagasaki 852-8501, Japan, the ¶Osaka University Faculty of Dentistry, Suita, Osaka 565-0871, Japan, the ||Molecular Biology Section, National Institutes of Health, NIDCR, Bethesda MD, 20892, and the **Department of Genetics and Development, College of Physicians and Surgeons, Columbia University, New York, New York 10032
Received for publication, December 20, 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-catenin signaling in nascent cartilaginous skeletal elements blocks chondrocyte hypertrophy and endochondral ossification, whereas signaling starting at later stages stimulates hypertrophy and ossification, indicating that Wnt/-catenin roles are developmentally regulated. To test this conclusion further, we created transgenic mice expressing a fusion mutant protein of -catenin and LEF (CA-LEF) in nascent chondrocytes. Transgenic mice had severe skeletal defects, particularly in limbs. Growth plates were totally disorganized, lacked maturing chondrocytes expressing Indian hedgehog and collagen X, and failed to undergo endochondral ossification. Interestingly, the transgenic cartilaginous elements were ill defined, intermingled with surrounding connective and vascular tissues, and even displayed abnormal joints. However, when activated -catenin mutant (
--catenin) was expressed in chondrocytes already engaged in maturation such as those present in chick limbs, chondrocyte maturation and bone formation were greatly enhanced. Differential responses to Wnt/-catenin signaling were confirmed in cultured chondrocytes. Activation in immature cells blocked maturation and actually de-stabilized their phenotype, as revealed by reduced expression of chondrocyte markers, abnormal cytoarchitecture, and loss of proteoglycan matrix. Activation in mature cells instead stimulated hypertrophy, matrix mineralization, and expression of terminal markers such as metalloprotease (MMP)-13 and vascular endothelial growth factor. Because proteoglycans are crucial for cartilage function, we tested possible mechanisms for matrix loss. --Catenin expression markedly increased expression of MMP-2, MMP-3, MMP-7, MMP-9, MT3-MMP, and ADAMTS5. In conclusion, Wnt/-catenin signaling regulates chondrocyte phenotype, maturation, and function in a developmentally regulated manner, and regulated action by this pathway is critical for growth plate organization, cartilage boundary definition, and endochondral ossification. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Wnt proteins are powerful secreted signaling factors that regulate a number of developmental processes (4, 5). The vertebrate Wnt family currently comprises 20 members. Wnt proteins act by binding to Frizzled and low density lipoprotein receptor-related protein cell surface receptors (4, 5). Upon Wnt binding, Frizzled receptors transduce signals via the -catenin-LEF/TCF pathway (5, 6), Ca2+-calmodulin-PKC pathway (7), or JNK-dependent pathway (8). The -catenin-LEF/TCF pathway is the best characterized to date and is conserved from Drosophila to humans (5, 6). In the absence of Wnt binding to Frizzled receptors, -catenin is found to form complexes with cytoplasmic proteins such as glycogen synthase kinase 3, Axin, and adenomatous polyposis coli, and remains phosphorylated (9, 10). Phosphorylated cytoplasmic -catenin molecules become ubiquitynated and are directed for proteasome-assisted degradation (11). On the other hand, when Wnt proteins bind to the receptors, degradation is inactivated and a "free" form of -catenin accumulates in the cytoplasm (9, 10, 12). Accumulated -catenin molecules translocate to the nucleus and serve as a co-activator of resident LEF/TCF transcriptional factors, resulting in complexes that bind to response sequences and modulate expression of target genes (6, 12).
Wnt proteins are involved in skeletogenesis. Seminal observations by Dealy and co-workers (13) first showed that Wnt-7a is expressed by early limb bud ectoderm and influences skeletal patterning along the dorsal-ventral axis. Wnt-3a was found to share similar expression patterns and possibly similar roles (14). Subsequent studies indicated that Wnt-1, -4, and -7a exert inhibitory influences on differentiation of mesenchymal cells into chondrocytes, whereas Wnt-5a and -5b stimulate it (15, 16). Wnt-5a and Wnt-5b were also shown to have strong influences on skeletogenesis in chick and mouse embryos (17–19). Of particular relevance to the present study are two recent reports on Wnt/-catenin roles in skeletogenesis using transgenic approaches (20, 21). Akiyama et al. (20) found that continuous expression of an activated -catenin form (--catenin) in nascent chondrocytes using collagen II promoter sequences blocks further chondrocyte development and also ossification, as indicated by low expression of chondrocyte master regulator Sox9 and bone regulator Runx2. A detailed analysis of the chondrocyte phenotype and maturation program was, however, not performed. Guo et al. (21) used a similar approach and found severe defects in long bone joint formation also. On the other hand, previous studies from our groups indicated that when Wnt/-catenin signaling is experimentally activated at later stages and specifically in chondrocytes already undergoing and advancing toward maturation, activation resulted in a marked stimulation of hypertrophy (22, 23). Together, the above studies lead to the intriguing and important notion that Wnt/-catenin signals may not have a single role in skeletogenesis, but their interventions in the regulation of this complex multistep process may be multiple and developmentally prescribed. The present study was carried out to test these hypotheses more directly using transgenic approaches in mice and gain-of-function studies in chick embryo limbs and chick chondrocytes in culture. The results provide support for our hypotheses and reveal also that the Wnt/-catenin signaling pathway has critical roles in maintenance of cartilage tissue integrity and formation of proper tissue boundaries.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-catenin fused to the COOH terminus of murine LEF-DNA (24), were cloned into the NotI site of a collagen 2 
1 (Col21)-based expression vector (25). Synthetic NarI site was introduced into a HindIII site of the above vector. The resulting vector contains the CA-LEF expression unit including the 5'-NarI site-Col21 promoter (nucleotide 1940–2971 of M65161
[GenBank]
), -globin intron cassette, CA-LEF, SV40 poly(A), Col21 enhancer (4930–5571 of M65161
[GenBank]
), NarI site-3' in pNASS backbone (Clontech). CA-LEF was tagged with HA antigen at the COOH-terminal site to monitor transgene expression. The expression unit of CA-LEF was excised by NarI digestion, purified, and injected into pronuclei of fertilized eggs from F1 hybrid mice (C57BL/6 X C3H). Transgenic embryos were identified by PCR using two pairs of primers: 5'-TGCAGCTTTATCCAGGCTGGTCAG-3' and 5'-CACCCATCTCATGCTCCATCATCATAGG-3', which amplify the 345-bp fragment of CA-LEF; 5'-TGGTCATCATCCTGCCTTTCTC-3' and 5'-ATTAGGTCACTGTCCGTGTGGG-3', which amplify the 256-bp fragment of part of the -globin intron and the LEF-1 DNA binding domain. Genomic DNA was prepared from liver tissue. Expression of the transgene at the protein level was analyzed by immunostaining for HA antigens. Seventy-two of 377 embryos carried CA-LEF transgene and 11 embryos expressed the transgene products. Skeletal, Histological, and Immunohistochemical Examinations— For anatomical examination, whole skeletons of E18.5 embryos were stained with alizarin red S and Alcian blue as described (25). For light microscopy, tissues from E18.5 embryos were fixed in 4% paraformaldehyde, 0.1 M phosphate buffer, and the resulting paraffin sections (6 µm in thickness) were stained with hematoxylin and eosin (H&E). For immunohistochemistry, sections were de-masked by treatment with 0.2% pepsin in 0.02 N HCl for 15 min at 37 °C, and were incubated with rabbit anti-mouse tenascin-C antibody (26) or anti-MMP-2 antibody (Chemicon International Inc., Temecula, CA) for 16 h at 4 °C. After rinsing, sections were incubated with biotinylated anti-rabbit IgG (Vector Laboratories, Inc., Burlingame, CA) and then with Cy3-conjugated streptavidin (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA); each incubation was for 30 min at room temperature.
In Situ Hybridization—Tissue sections using in situ hybridization was carried out with digoxigenin-conjugated or 35S-labled riboprobes (25, 27). cDNA fragments of mouse Type IX (317–805 of NM_007740 [GenBank] ) and Type X (1302–1816 of NM_009925 [GenBank] ) collagens, aggrecan (880–1733 of NM_007424 [GenBank] ), Sox9 (569–1520 of AF421878 [GenBank] ), IHH (897–1954 of NM_010544 [GenBank] ), Gas1 (1559–2452 of NM_008086 [GenBank] ), GDF5 (1585–2172 of U08337 [GenBank] ), ChM-1 (156–1195 of U43509 [GenBank] ), and ERG (1007–1029 of AB07380) were amplified by reverse transcriptase (RT)-PCR from E15 mouse limb bud cDNA, subcloned into pGEM-T vector (Promega Corp., Madison, WI), and used as templates to generate RNA probes. Chick CD44 (438–915 of NM_204980 [GenBank] ) and Gli3 (688–1131 of AF0022818) were amplified from E8 chick wing bud cDNA by RT-PCR and subcloned into pGEM-T vector.
Cell Cultures and Viral Infection—Chondrocytes isolated from the caudal (for immature cells) and cephalic (for mature cells) portions of day 17 chick embryo sterna (SPAFAS, c/o) were cultured in high-glucose Dulbecco's modified Eagle's medium containing 10% fetal bovine serum (28). The cells were infected with concentrated RCAS virus encoding the active form of --catenin that lacks the amino-terminal domain (-catenin) (14) or chick Wnt-14 (29) (kindly provided by Dr. C. J. Tabin, Harvard University) and subcultured at the density of about 1.0–1.5 x 104/cm2 by trypsinization after 1 week, and maintained in complete medium containing 10 µg/ml of ascorbic acid. By passage one, more than 85% of the cells were routinely infected as revealed by immunocytochemistry of viral antigens (30); expression of introduced genes was confirmed by Western blot. For matrix mineralization, cultures were provided with 1 mM -glycerophosphate (31). Proteoglycan accumulation and matrix mineralization in the cell layer were visualized by histochemistry (30).
Freshly isolated immature and mature chondrocytes were cultured until subconfluent and were then transfected with Ax2-Luc promoter reporter vector that contains Axin-2 promoter, exon 1, and intron 1 (32) in the absence or presence of the pUSE expression vector encoding --catenin or the dominant negative form of LEF (that lacks amino acids 17–264 of murine LEF-1) (24). Transfections were carried out with TransIT-LT1 transfection reagent (Mirus Bio Corp., Madison, WI) according to the manufacturer's instructions. Luciferase activity was measured 2 days after transfection by standard methods.
RNA Isolation and RT-PCR—Total RNA was isolated by the guanidine isothiocyanate method (31). First strand cDNA was synthesized from 1 µg of total RNA with 1 µM of random 9-mer primer (PerkinElmer Life Sciences) using SuperScript IITM reverse transcriptase (Invitrogen) at 42 °C for 45 min. Subsequent amplification was performed with Premix Taq (TaKaRa EX Taq version, Takara Mirus Bio Inc., Madison, WI) for 25–35 cycles under the following conditions: 94 °C for 30 s, 58–60 °C for 1 min, and 72 °C for 1 min. Primer sequences for RT-PCR amplification are follows: 5'-TTCTGCTCACTTGGATCCACTG-3' and 5'-CGATCTCCATCGTGACGTTGTA-3' for 13–324 of vascular endothelial growth factor (AB011078 [GenBank] ), 5'-GCAACAGAAACGAGGAGCAAAGCAT-3' and 5'-CAGGGGAAGAAGTAGGTTCACCG TC-3' for 438–915 of CD44 (NM_204980 [GenBank] ), 5'-AGCGGATGCTCAAGGGGTTCTATG-3' and 5'-GGTGCAGATGGCACATTTGATGAG-3' for 1814–2409 of chordin (NM_204980 [GenBank] ), 5'-TGTCTATCTCCCGCTGTCTGATGC-3' and 5'-CGACTTCGGTCTTGACTTTGGTGCG-3' for 688–1131 of Gli3 (AF0022818), 5'-TCCTTTGCTTCTGGATTTCACGGTG-3' and 5'-CTCTGGTCTTTCTGTGGGTGCCATT-3' for 523–960 of MMP3 (XM_417175 [GenBank] ), 5'-TGCAACAGCCCAGATGTGGAATCC-3' and 5'-CCTCGCCTATCATCCCAATGAAGTG-3' for 295–805 of MMP7 (Ensemble data base ID: ENSGALT00000027770), 5'-GCCCTATGAAGACATCCGCCAGAA-3' and 5'-TGGAAGTTCTCCGTGTCCATCCACT-3' for 84–407 of MT2-MMP (ENSGALT 00000001489), 5'-TCACTCCCCTCACCTTCACCCAGAT-3' and 5'-ACCCATCGCTGTTGCCACCATTG-3' for 452–856 of MMP9 (AF222690 [GenBank] ), 5'-GCCTTGTGCCTCCTTGTCTTGGTT-3' and 5'-AAAGTTCCACCGTGAGTTGTTGCGT-3' for 1837–2365 of cMT3-MMP (NM_205197 [GenBank] ), 5'-GGAGGCGTTCAGTTTGCTTATCGTC-3' and 5'-TCCACCACACACCCCACATTTATCATA-3' for 832–1257 of ADAMTS5 (ENSGALT00000025470). The resulting products were sequenced to confirm specific amplification of the target transcripts. All other PCR primers were described previously (22, 33).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-catenin (CA-LEF) was expressed in nascent chondrocytes in developing mouse embryo skeletal elements, using cartilage characteristic collagen II promoter/enhancer sequences to drive the transgene (25). CA-LEF is a fusion protein consisting of the DNA binding site of mouse LEF-1 and the transactivation domain of -catenin, and has activity in chondrocytes similar to --catenin used in previous studies (24). Transgenic mice expressing CA-LEF died at birth of respiratory failure and displayed severe dwarfism with shortened limbs and a flattened skull (Fig. 1B) compared with wild type littermates (Fig. 1A). Skeletal staining with alizarin red and Alcian blue revealed that mineralization was markedly reduced in trunk and limbs and affected ribs, vertebrae, and long bones (Fig. 1, A and B). Under-mineralization was also evident in the cranial base (Fig. 1D, arrow) compared with wild type (Fig. 1C), but calvaria mineralization appeared largely unaffected (Fig. 1, A and B). In addition to mineralization defects, transgenic skeletons displayed abnormalities in overall patterning, development, and growth, particularly in the limbs. Transgenic scapula, humerus, radius, and ulna were considerably short (Fig. 1H) and hind limb elements were sharply hypoplastic and totally un-mineralized (Fig. 1J) compared with corresponding wild type elements (Fig. 1, G and I). Ulna and radius had undergone fusion in some embryos (Fig. 1H, arrow) and digit joints were poorly formed (Fig. 1, H and J, arrowheads). Interestingly, the extent and overall patterns of Alcian blue staining suggested that transgenic cartilages and their extracellular matrices were actually not uniform and homogenous (Fig. 1L, arrows), although wild type cartilage displayed typical seamless matrix homogeneity (Fig. 1K).
Growth Plate Defects—Histologically, long bone elements in wild type E18.5 limbs had a typical organization consisting of a well formed articular cap and joints, well defined zones of chondrocyte maturation in the growth plates, and active endochondral ossification in the diaphysis (Figs. 2, A, C, E, and G). In contrast, transgenic elements were completely disorganized (Fig. 2, B, D, F, and H). There was no clear histological demarcation of articular cap and joints, and typical growth plates were absent. Most if not all the chondrocytes, regardless of their localization with a given element, displayed uniform, round and relatively small-sized morphology and shape (Fig. 2, D and H, arrows), thus differing sharply from wild type chondrocytes displaying their characteristic and marked changes in cytoarchitecture and phenotype in different growth plate zones (Fig. 2, C and G). Interestingly, transgenic cartilage was not clearly distinguishable from adjacent tissues, and there was common occurrence of an apparent inter-mixing of cartilage tissues with adjacent connective and vascular tissues (Fig. 2, D and H, arrowheads). The latter probably accounts for the lack of uniform Alcian blue staining seen anatomically in similar transgenic specimens above (Fig. 1L).
To verify that the above histological defects were associated with expression of CA-LEF in cartilage, sections were processed for immunohistochemistry using antibodies to HA tag present in the CA-LEF-encoding transgene. Transgenic cartilages were clearly stained by the HA antibodies, although surrounding tissues were not (Fig. 2J). In addition, every tissue in the wild type specimens was negative (Fig. 2I), reaffirming specificity of expression and staining.
Gene Expression Patterns—The severe histological defects and chondrocyte cellular uniformity seen in transgenic growth plate cartilage suggested that the maturation program of chondrocytes had been markedly affected by CA-LEF expression. To characterize such possible changes, longitudinal sections of E18.5 limb skeletal elements were processed for histochemical staining with Safranin O to depict matrix proteoglycan deposition and for in situ hybridization to define the molecular cell phenotypes. Wild type specimens stained strongly with Safranin O both in their articular cap and growth plate, and low staining was seen in subadjacent endochondral bone and surrounding connective tissues as expected (Fig. 3A). In contrast, transgenic cartilages stained in a dull and substandard manner, and staining of cartilaginous masses was often not uniform (Fig. 3B, arrows). In situ hybridization showed that wild type articular cap and growth plate chondrocytes strongly expressed genes encoding cartilage-characteristic matrix macromolecules, such as Collagen IX and Aggrecan, and showed a prescribed reduction in expression of these genes in the hypertrophic zone (Fig. 3, C and E). The cartilage master regulator Sox9 displayed an identical and typical pattern (Fig. 3G), and the important signaling factor Indian hedgehog (IHH) was expressed in the pre-hypertrophic zone (Fig. 3I). Transgenic cartilage displayed very low and scattered expression of collagen IX and aggrecan (Fig. 3, D and F), poor expression of Sox9 (Fig. 3H), and undetectable expression of IHH (Fig. 3J). Not surprisingly then, expression of genes associated with chondrocyte hypertrophy and endochondral ossification, which is Collagen X, was completely absent in transgenic cartilage (Fig. 3L), whereas the gene exhibited a strong and restricted pattern in wild type (Fig. 3K).
|
|
|
|
Blood Vessel Analysis—To further examine the mechanisms leading to tissue boundary alterations at epiphyseal ends and along the shaft of long bone elements, we processed longitudinal sections for immunostaining with antibodies to tenascin-C, a protein characteristic of developing articular cartilage and present at tissue-tissue interfaces also (25, 36). Wild type E18.5 limb sections displayed strong and characteristic tenascin-C staining, with the protein decorating the epiphyseal articular layer and metaphyseal perichondrium (Fig. 5A). In transgenic sections, however, the protein was much more abundant and widely distributed in cartilage and surrounding tissues (Fig. 5, B and F), a clear sign of chaotic tissue-tissue interfaces and relationships. Thus, we next examined the distribution of blood vessels, using antibodies to the cell surface marker CD34 (which is expressed by mesenchymal and vascular endothelial cells) and CD31 (which is restricted to endothelial cells). In wild type sections, CD34- or CD-31-positive cells were present in periarticular tissue, perichondrium, and periosteum (Fig. 5, G and J, arrows) and were absent from articular cap cartilage itself and most of the growth plate cartilage (Fig. 5, G and J). There was, however, clear and well defined staining at the chondro-osseus boundary reflecting ongoing vascular invasion and endochondral bone and marrow development (Fig. 5, G and J, arrowheads). In transgenic specimens, CD34- or CD-31-positive cells were distributed in a broader, ill defined and scattered manner (Fig. 5, H, I, K, and L). Some of the positive cells were in very close proximity and essentially within the cartilaginous tissue, appearing to be in the process of actively entering it (Fig. 5, H and K, arrows). Given the nature of CD-31 and CD-34 markers, the positive cells intermingling with cartilage likely represented a mixture of vascular and mesenchymal cells.
We asked then whether abnormal vascular distribution of transgenic cartilage is associated with, and may be because of, changes in expression of genes regulating angiogenesis. ChM-I is known as an angiogenic inhibitor and was originally identified in bovine cartilage (37). In wild type sections, ChM-1 transcripts were abundant and present from epiphyseal articular cartilage through much of the underlying growth plate, with an obvious reduction in the hypertrophic zone (Fig. 6A). ChM-1 expression was significantly reduced in transgenic specimens (Fig. 6B). We also examined the distribution of MMP-2, a protease responsible for extracellular matrix degradation during vascular formation (38). MMP-2 staining was clear and prominent in pre-hypertrophic and hypertrophic zones of wild type growth plates (Fig. 6, C and E), but was much more prominent and actually widespread in transgenic tissue and was not restricted to cartilage only (Fig. 6, D and F).
Chick Limb Studies—The transgenic data above indicate that continuous expression of CA-LEF in nascent chondrocytes driven by collagen II gene promoter sequences blocks further chondrocyte maturation and hypertrophy in transgenic mice, whereas our previous studies suggested that CA-LEF expression in cultured growth plate chondrocytes (already engaged in and pursuing maturation) accelerated their hypertrophy. The data raised the possibility that chondrocyte responses to Wnt/-catenin signaling activation depend on the cell developmental stage/status at the moment of activation. To test this possibility, we misexpressed a constitutive-active form of -catenin (--catenin) in day 5.5 chick embryo wing cartilaginous skeletal anlagen; at this stage of embryogenesis, humerus, radius, and ulna contain: diaphyseal growth plates with chondrocytes already engaged in maturation; and epiphyseal caps containing resting/pre-articular chondrocytes. We reasoned that simultaneous --catenin misexpression in diaphyseal and epiphyseal chondrocytes may elicit different responses and have different developmental consequences in diaphysis versus epiphysis.
Retroviral RCAS particles encoding --catenin were microinjected in the vicinity of cartilaginous anlagen in day 5.5 chick embryo limbs in ovo, and embryos were monitored and examined over time. Longitudinal sections of control day 11 embryos revealed that ulna had a typical elongated shape with an epiphyseal articular cap (Fig. 7C), a prominent growth plate extending from metaphysis to diaphysis (Fig. 7A), and a small amount of invading bone/marrow cell population in the diaphyseal center (Fig. 7, A and D, arrow). In --catenin-expressing ulna, however, there was extensive ossification and marrow deposition in the diaphysis (Fig. 7, B and F). However, the epiphyseal region had not undergone ossification and marrow deposition, and was instead composed of seemingly immature cartilage with ill-defined boundaries and weakly staining with Safranin O (Fig. 7E, arrowheads). Joint markers genes CD44 (39) and Gli3 (40) were widely expressed (Fig. 7, I and J), whereas they were confined to developing articular layers in control (Fig. 7, G and H). Immunostaining with -catenin antibodies (that recognize --catenin as well) showed that the protein was widely distributed in epiphyseal and diaphyseal cartilage and adjacent tissues in virally infected specimens (Fig. 7, M and N), whereas it was much less abundant and restricted in distribution in control tissues as expected (Fig. 7, K and L).
|
-catenin signal activation are developmentally regulated, populations rich in immature or mature chondrocytes isolated from caudal and cephalic sternal regions (31) were infected with --catenin-encoding RCAS viruses chondrocytes and were monitored over time in culture. Companion control cultures infected with insert-less viruses behaved as expected over time; the caudal cells maintained an immature and differentiated phenotype (Fig. 8A), whereas the cephalic cells displayed a hypertrophic phenotype (Fig. 8D). The responses of the two cell populations to --catenin overexpression differed markedly. The immature cells became fibroblastic and flat-shaped (Fig. 8B), but the mature cells remained differentiated (Fig. 8E). Similar responses were seen when jointassociated Wnt-14 (29) was overexpressed in place of --catenin (Fig. 8, C and F). Histochemical staining showed that --catenin overexpression reduced proteoglycan content and accumulation in both immature and mature cultures (Fig. 8G, AB), but stimulated matrix calcification in mature cultures only (Fig. 8G, AR). Again, Wnt-14 overexpression elicited similar responses (Fig. 8G).
To further characterize the cell phenotype, we determined expression of various chondrocyte markers by RT-PCR. Overexpression of --catenin or Wnt-14 inhibited aggrecan and type IX collagen gene expression in both immature and mature chondrocytes (Fig. 9, A and B, respectively). However, it stimulated markers of terminal chondrocyte maturation (MMP-13 and vascular endothelial growth factor) in mature cultures (Fig. 9D) consistent with data on matrix calcification (Fig. 8G, AR), but had no major effect in immature cultures (Fig. 9C). Type X collagen expression, which normally decreases in the mineralizing zone of growth plate, was decreased by --catenin or Wnt-14 (Fig. 9D).
|
--catenin and Wnt-14 are able to activate the Wnt/-catenin pathway in immature and mature chondrocytes equally well, control cultures and cultures overexpressing --catenin or Wnt-14 were transfected with a promoter reporter construct for Axin-2, a direct target of that pathway (32). Basic activity of the Axin-2 reporter was slightly higher in control mature than immature chondrocytes (Fig. 8H), a likely reflection of endogenous differences in ongoing signaling. Reporter activity was enhanced about 5-fold by --catenin and about 3-fold by Wnt-14 in immature and mature cultures, indicating that the two populations did possess similar Wnt/-catenin signal activation capacities. This was confirmed by similar responses to overexpression of Wnt-3a or Wnt-8 (Fig. 8H), both of which are well known activators of the Wnt/-catenin pathway. Increases in reporter activity were all counteracted by co-expression of dominant-negative LEF (Fig. 8H), re-affirming specificity of responses.
MMP Gene Expression—The extracellular matrix is crucial for cartilage structure and function, and thus our data indicating a major loss of proteoglycan content and accumulation in cultures overexpressing --catenin or Wnt-14 was of particular interest. Thus, we asked whether this response is accompanied by increases in gene expression for metalloproteases, enzymes intimately linked to matrix turnover, and remodeling (41–43). Indeed, --catenin or Wnt-14 overexpression led to sharp increases in expression of MMP-2,-7, and -9, MT3-MMP, and ADAMTS5 in both immature and mature chondrocyte cultures (Fig. 9, E and F). MMP-3 expression was stimulated in mature chondrocytes only (Fig. 9F), whereas expression of MT-2-MMP was increased in mature cultures but decreased in immature cultures (Fig. 9, E and F).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
--catenin is overexpressed in the developing chick embryo limb long bone anlagen, endochondral ossification and bone deposition are enhanced in diaphyseal regions, whereas immature but disorganized cartilage persists at the epiphyseal regions. The data strongly indicate that the responses of chondrocytes to Wnt/-catenin signaling and the phenotypic and functional consequences are strictly dependent on stage/status of the cells at the moment of signal activation. This is reaffirmed by the very distinct responses to signal activation seen in cultures of chick immature and mature chondrocytes. The data correlate quite well with our previous data showing that whereas much of endogenous -catenin is cytoplasmic in proliferative and pre-hypertrophic zones of the growth plate, it shifts to a nuclear localization in hypertrophic chondrocytes (22). It is clear then that Wnt/-catenin signals regulate chondrocyte behavior and function in a developmentally regulated and maturation-dependent manner and that tight control of this signaling pathway is required for normal chondrocyte behavior and maturation and normal skeletogenesis. Chondrocyte De-stabilization and Tissue Boundary Loss—A prominent and obvious feature of transgenic CA-LEF-expressing cartilaginous elements is that their chondrocytes are relatively uniform in size and shape, do not display the changes in cytoarchitecture and volume characteristic of normal growth plates, and are tightly packed together, probably a result of suboptimal deposition of extracellular matrix. The tissue itself is not homogenous and not well separated from adjacent tissues. These histological and cellular alterations resemble those seen in thanatophoric dysplasia or achondroplasia (44). In terms of gene expression, the transgenic chondrocytes display poor expression of such typical genes as Aggrecan, Collagen IX, and Sox9, and virtually undetectable expression of IHH and Collagen X. These data indicate that the transgenic chondrocytes were not only unable to mature and become hypertrophic, but were also likely to be undergoing phenotypic de-stabilization. The latter is suggested by the poor expression of Sox9, a transcription factor that acts as a master regulator of chondrocyte phenotype (45), a change also documented by Akiyama et al. (20).
In addition to poor maturation, chondrocytes in CA-LEF transgenic mice express joint marker genes widely and ectopically, including GDF5, GAS1, and ERG. These genes are thought to coordinate multiple events needed for joint formation, including formation of interzone, organization of joint structures, and maintenance of joint function (27, 34, 35, 46). Our findings are consistent and agree well with recent reports in which Wnt-14 and -catenin signaling were found to induce ectopic joint marker expression (21, 29). These authors interpreted the findings to indicate that Wnt-14 and -catenin are not only involved in joint formation, but may actually be sufficient to induce it. This conclusion, however, may be premature. In addition to inducing ectopic joint markers, overexpressed Wnt-14 and active -catenin actually disturbed joint, particularly in the digits where the skeletal elements were fused and the chondrocytes were uniform and packed together. We observed a similar disturbance of joint formation in CA-LEF-expressing transgenic cartilaginous limb elements, but did not detect de novo joint formation at ectopic sites. Thus, whereas Wnt/-catenin signaling pathway is certainly important for joint formation, it may be insufficient to induce it and may depend on upstream mechanisms dictating site and timing of joint formation initiation.
|
Transgenic cartilage invasion might mimic pathological situations such as those seen in joint degenerative disorders, which exhibit sever loss of cartilage matrix and tissue by itself and invasion by the surrounding tissues. In the present study we demonstrated that --catenin overexpression strongly stimulates expression of matrix degradation enzymes such as MMPs and MT-MMPs in chondrocytes. We found also that Wnt/-catenin signal stimulates expression of MMPs and aggrecanases in mouse chondrogenic cells.2 In osteoarthritis and rheumatoid arthritis, up-regulation of a variety of matrix degradation enzymes including MMPs and aggrecanases are crucial and considered responsible for cartilage matrix loss (42, 50). Interestingly, we found that osteoarthritic cartilage displays signs of Wnt/-catenin activation.3 Other investigators have reported that Wnt-1, an activator of the -catenin signal, and other Wnts are up-regulated in synovial cells in rheumatoid arthritis (51), and that functional variants within sFRP3, which is an endogenous antagonist of the Wnt/-catenin signal, are associated with hip osteoarthritis in females (52). Thus Wnt/-catenin signals may activate cartilage matrix catabolism and may have roles in cartilage destruction under pathological conditions.
|
Control of Wnt/-Catenin Signaling and Action—The data described here and previous reports by others make it quite clear that restriction, activation, and modulation of Wnt/-catenin signaling are crucial for chondrogenesis and skeletogenesis (20–23). The pathway must be kept relatively inactive for condensed mesenchymal cells to differentiate into chondrocytes and for chondrocytes to be able to assemble into well defined skeletal anlagen with characteristic growth plates. The pathway, however, must be activated to allow maturing chondrocytes to complete their development into hypertrophic mineralizing cells, which in turn orchestrate their replacement with endochondral bone and marrow. What could then be possible mechanisms activating and modulating Wnt/-catenin signaling in developing chondrocytes? It should be pointed out that a number of Wnt proteins, including Wnt-2b, Wnt-3a, Wnt-4, Wnt-7b, and Wnt-8, are expressed by developing chondrocytes or neighboring cells in vivo (14, 18, 22, 53, 54) and that these proteins all have the ability to activate Wnt/-catenin signaling (53–57). In addition, there are well known extracellular components that can dampen the degree of signaling by affecting Wnt binding to their receptors (58). These components include soluble Frps, Dickkopf (DKK), and WIF. Interestingly, onset of Frp3/Frzb-1 expression coincides with onset of chondrogenesis (59, 60), but expression is down-regulated in the hypertrophic zone of the growth plate, and Frzb-1 neutralizes the inhibition of chondrogenesis by misexpression of Wnt-8 (22). Thus, it is possible that continuous Frzb-1 expression in developing chondrocytes may contribute to prevent Wnt action and may maintain the cells in a healthy and functional state. The cells would be able to form well defined cartilaginous anlagen, organize growth plates, and establish well defined tissue boundaries. Frzb-1 down-regulation in the hypertrophic zone, however, would allow strong Wnt action, permitting the cells to terminally differentiate and favoring the normal transition from hypertrophic cartilage to bone.
|
-catenin signaling pathway actions and roles is Wnt-5a. This Wnt is first expressed in cartilage anlagen and expression shifts to perichondrium with further development (16, 19). Deletion of the Wnt-5a gene in mice causes a phenotype quite similar to that seen in CA-LEF transgenic mice here (17). Wnt-5a-null mouse embryos are dwarfs with shortened and deformed limbs, and display cartilaginous elements with low expression of Sox9, IHH, PTH/PTHrp receptor, and Type X collagen and no endochondral ossification (17). Wnt-5a has been shown to inhibit the Wnt/-catenin pathway through the calmodulin-dependent protein kinase II pathway or an unknown pathway that inhibits -catenin degradation (61). These findings suggest that Wnt-5a could serve as a strong inhibitor of Wnt/-catenin signaling during cartilage development. Indeed the limb buds of Wnt-5a-null mouse embryos display ectopic up-regulation of Wnt/-catenin as revealed by promoter reporter activity (61). Thus, Wnt-5a may be part of the normal mechanisms involving also Frzb-1 by which Wnt/-catenin signaling is maintained relatively low at the initial stages of skeletogenesis and cartilage development, and is then markedly up-regulated in hypertrophic mineralizing cartilage.
Last, it is important to discuss Sox9. The factor is continuously expressed starting with pre-cartilaginous mesenchymal cells up to pre-hypertrophic chondrocytes, and is then downregulated in hypertrophic cells (62). Sox9 inhibits Wnt/-catenin signals by competing with -catenin binding to LEF/TCF proteins, and experimental ablation of Sox9 in cartilage causes a phenotype similar to that following chronic activation of the Wnt/-catenin pathway (20). Because Sox9 expression precedes differentiation of mesenchymal cells into chondrocytes, this molecule also could antagonize the Wnt/-catenin signal from onset of cartilage formation up until the pre-hypertrophic stage. Terminal hypertrophic chondrocyte maturation would thus entail a down-regulation of Sox9, in coordination with down-regulations in Wnt-5a and Frzb1 and concurrent boost in Wnt/-catenin signaling.
|
FOOTNOTES |
|---|
To whom correspondence should be addressed: Dept. of Orthopedic Surgery, Jefferson Medical College, Thomas Jefferson University, 1015 Walnut St., Curtis Bldg. Suite 501, Philadelphia, PA 19107. Tel.: 215-955-7624; Fax: 215-955-9159; E-mail: Motomi.Iwamoto{at}jefferson.edu.
1 The abbreviations used are: MMP, metalloprotease; IHH, Indian hedgehog; HA, hemagglutinin; RT, reverse transcriptase; CA-LEF, constitutive active form of mouse LEF-1.
2 Y. Tamamura, T. Otani, M. Iwamoto, and M. Enomoto-Iwamoto, manuscript in preparation.
3 T. Otani, M. Iwamoto, and M. Enomoto-Iwamoto, submitted for publication.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
